Heatmap-Guided Selective Feature Attention for Robust Cascaded Face Alignment
Face alignment methods have been actively studied using coordinate and heatmap regression tasks. Although these regression tasks have the same objective for facial landmark detection, each task requires different valid feature maps. Therefore, it is not easy to simultaneously train two kinds of task...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-05-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/10/4731 |
| _version_ | 1797598385013784576 |
|---|---|
| author | Jaehyun So Youngjoon Han |
| author_facet | Jaehyun So Youngjoon Han |
| author_sort | Jaehyun So |
| collection | DOAJ |
| description | Face alignment methods have been actively studied using coordinate and heatmap regression tasks. Although these regression tasks have the same objective for facial landmark detection, each task requires different valid feature maps. Therefore, it is not easy to simultaneously train two kinds of tasks with a multi-task learning network structure. Some studies have proposed multi-task learning networks with two kinds of tasks, but they do not suggest an efficient network that can train them simultaneously because of the shared noisy feature maps. In this paper, we propose a heatmap-guided selective feature attention for robust cascaded face alignment based on multi-task learning, which improves the performance of face alignment by efficiently training coordinate regression and heatmap regression. The proposed network improves the performance of face alignment by selecting valid feature maps for heatmap and coordinate regression and using the background propagation connection for tasks. This study also uses a refinement strategy that detects global landmarks through a heatmap regression task and then localizes landmarks through cascaded coordinate regression tasks. To evaluate the proposed network, we tested it on the 300W, AFLW, COFW, and WFLW datasets and obtained results that outperformed other state-of-the-art networks. |
| first_indexed | 2024-03-11T03:20:25Z |
| format | Article |
| id | doaj.art-7307c938d9d24e1a8a2a2226068d10cc |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T03:20:25Z |
| publishDate | 2023-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-7307c938d9d24e1a8a2a2226068d10cc2023-11-18T03:11:44ZengMDPI AGSensors1424-82202023-05-012310473110.3390/s23104731Heatmap-Guided Selective Feature Attention for Robust Cascaded Face AlignmentJaehyun So0Youngjoon Han1Department of Electronic Engineering, Soongsil University, Seoul 06978, Republic of KoreaSchool of AI Convergence, Soongsil University, Seoul 06978, Republic of KoreaFace alignment methods have been actively studied using coordinate and heatmap regression tasks. Although these regression tasks have the same objective for facial landmark detection, each task requires different valid feature maps. Therefore, it is not easy to simultaneously train two kinds of tasks with a multi-task learning network structure. Some studies have proposed multi-task learning networks with two kinds of tasks, but they do not suggest an efficient network that can train them simultaneously because of the shared noisy feature maps. In this paper, we propose a heatmap-guided selective feature attention for robust cascaded face alignment based on multi-task learning, which improves the performance of face alignment by efficiently training coordinate regression and heatmap regression. The proposed network improves the performance of face alignment by selecting valid feature maps for heatmap and coordinate regression and using the background propagation connection for tasks. This study also uses a refinement strategy that detects global landmarks through a heatmap regression task and then localizes landmarks through cascaded coordinate regression tasks. To evaluate the proposed network, we tested it on the 300W, AFLW, COFW, and WFLW datasets and obtained results that outperformed other state-of-the-art networks.https://www.mdpi.com/1424-8220/23/10/4731face alignmentfeature attentionheatmap regressioncoordinate regressionmulti-task learning |
| spellingShingle | Jaehyun So Youngjoon Han Heatmap-Guided Selective Feature Attention for Robust Cascaded Face Alignment Sensors face alignment feature attention heatmap regression coordinate regression multi-task learning |
| title | Heatmap-Guided Selective Feature Attention for Robust Cascaded Face Alignment |
| title_full | Heatmap-Guided Selective Feature Attention for Robust Cascaded Face Alignment |
| title_fullStr | Heatmap-Guided Selective Feature Attention for Robust Cascaded Face Alignment |
| title_full_unstemmed | Heatmap-Guided Selective Feature Attention for Robust Cascaded Face Alignment |
| title_short | Heatmap-Guided Selective Feature Attention for Robust Cascaded Face Alignment |
| title_sort | heatmap guided selective feature attention for robust cascaded face alignment |
| topic | face alignment feature attention heatmap regression coordinate regression multi-task learning |
| url | https://www.mdpi.com/1424-8220/23/10/4731 |
| work_keys_str_mv | AT jaehyunso heatmapguidedselectivefeatureattentionforrobustcascadedfacealignment AT youngjoonhan heatmapguidedselectivefeatureattentionforrobustcascadedfacealignment |